Episode 33: Why People Analytics Needs to Behave Like a Startup (Interview with Dirk Jonker, CEO of Crunchr)

The field of people analytics has made significant progress in the last five years and was recently recognised by LinkedIn in its global talent trend report as one of four key trends reshaping the way organisations recruit and retain its people. Whilst people analytics has travelled far it still has a long way to go. Challenges abound on topics like ethics, upskilling HR and scaling people analytics across the organisation.

My guest on this week’s show agrees, as founder and CEO of Crunchr, Dirk Jonker has been working in the people analytics field for more than a decade and has helped numerous clients to build, operationalise and scale people analytics in their companies.

You can listen below or by visiting the podcast website here.

In our conversation Dirk and I will discuss:

• Why people analytics needs to behave more like a startup

• How to bring people analytics to scale

• Some of the key considerations for organisations that are looking to build their own technology to aggregate and visualise people data

• Why many organisation's approach to up-skilling HR is doomed to failure

• Whether AI and automation is a threat or an opportunity for HR

This episode is a must listen for anyone in a workforce or people analytics role, those involved in buying or developing HR technology in their organisations and any HR or business professional seeking to understand more about the people analytics field.

Support for this podcast is brought to you by Crunchr. To learn more, visit https://www.crunchrapps.com.

Interview Transcript 

David Green: Today I am delighted to welcome Dirk Jonker, the founder and CEO of Crunchr to The Digital HR Leaders Podcast. Great to have you Dirk.

Dirk Jonker: Thanks for having me.

David Green: Would you just like to give listeners a quick introduction to yourself and your background.

Dirk Jonker: Sure. So I'm Dirk. I am super fanatic about people analytics and that started for me 15 years ago.

I used to work as a consulting actuary and most of my colleagues ended up in insurance companies or retirement funds. But if you really look at the core of what an actuary does, it is basically translating uncertain business scenarios into data and insights today to make really big decisions.

That was really what I loved. I started to become really curious to see if you could use the same elegant mathematics on the data that companies already sit on to give HR the opportunity to influence its success way beyond their HR function. So basically, this was my dream, I took two suitcases went off to New York with a one way ticket and learned from the best people how to basically do business restructuring and business optimisation. I constantly saw the link between the people data, the business data, elegant mathematics and how can we use this to make impact.

On the personal side of that I met my wife there, got twins and then moved back to Amsterdam where I am originally from, as you can hear by the accent.

David Green:  So I think we probably both agree that doing analytics on people data is probably a bit more interesting than doing it on insurance data and financial data.

Dirk Jonker: Yeah. I mean, the concepts are the same but I really find a lot of joy in using exactly the same type of mathematics and using it on different new fields.

David Green: As you said we have both been in the space for quite a while and my sense is that people analytics has come a long way, but as we were talking about when we were prepping for this last week, we both feel it has got a long way to go. What is your view on the state of people analytics today?

Dirk Jonker: I totally agree. We have come a long way, every week there is a HR analytics conference somewhere in the world. At universities you can do now a fantastic education and we have seen fantastic examples in the world. Also thanks to you as a rainmaker in HR analytics, but I also feel that we are getting a bit stuck, we jump from hype to hype. What we just talked about, the HR tech conferences,is basically what we have been doing for the last seven years.

Seven years ago it was all about, hey, what do you do? This sounds interesting. The next year it was about, hey, how do you go beyond reporting? What kind of dash boarding do you provide? Then suddenly it was the age of AI. What do you do with machine learning? Now it is all about ONA. So we jumped from hype to hype but basically we forgot one important thing and that is to show return on investments.

David Green: Yeah. Yeah. I think, we are certainly seeing in some of the work that we are doing, that more and more organisations as they reach a certain level of maturity, actually start thinking, how do we turn some of this work that we are doing and actually how do we get tangible business outcomes from it. I think if there is a big place for people analytics to learn it is probably there I think.

Dirk Jonker: Yes totally agree. I mean, don't get me wrong, the whole projects that everybody did are super useful because it showed the potential. But maybe we can make a comparison to startups, how they are built. They also have an idea, they do some prototyping and then there comes a moment when they decide either we are going to do this for real or close the project and move on to something else. We pivot. So I feel that this is really the time for HR to say, thank you very much HR leadership for the seed investment. We built a team of five, six people, we had the opportunity to do fantastic projects, we demonstrated basic potential, but now it is time to reconsider how the target indirection model is going to be.

That is important and that is really what I see companies struggling with.

David Green: I think we all know there are companies out there that are kind of leading the way on people analytics and maybe what they have done differently to those that are still, as you said around five or six people, is they have actually demonstrated that value. They have got more investment and they demonstrate even more value. So we see companies obviously like Microsoft, like IBM, some of the big banks who have got reasonably large teams now and they are able to scale what they are effectively doing. So as you said, it is kind of moving from that kind of project mentality to a product mentality.

Dirk Jonker: Yes exactly. But even very big financial institutions run into challenges. Let's compare this to building a product because that is basically what they do, you go from project to product to really get skill in your organisation. So they get investments because they focus on very interesting projects that generate business value. But then the question is basically how to really scale, because even though you have 10 or 15 or maybe even 20 people, the impact that you could make without proper product or technology is just those 15 people times eight hours in a day times 260 working days in a year. So the question is how to scale this to the broad field HR to make this accessible and you will see that even large financial organisations are struggling.

Why? Maybe compare this to wobbling stones, we are at the beach, we both have kids, the thing that kids do is they take stones and put another stone and put another stone and another stone and they make this beautiful architectural structure. This is what HR does exactly the same.

They have payroll, another payroll, another payroll, they put a Workday or Successfactors or whatever on top, then they feel that they need some kind of reporting mechanism because they want to generate value. So they put a data extraction layer. Then they put a database, then they put another tool like a BI tool to generate insights. Then there is another tool on top, which basically helps to analyse this data a bit deeper, so from dashboarding to analysis. This starts to wobble a bit and the more pressure we put on this top stone the more people will land in people analytics, the more people we give access to this the more interesting the data will get. It starts to breakdown and this is exactly what I have seen in the last three months with two very major firms. So it is time to reconsider and rethink how we do people analytics.

David Green: Okay. What would be your tips around how you would rethink doing people analytics?

Dirk Jonker: All right so the first thing, Mercer wrote a beautiful article which says TIM before TOM.

And it is Targeted Interaction Model before Target Operating Model. So why is this? First we need to rethink on how and who are going to use the people insights that people analytics generates three or four years from now. To me it is basically the managers, why managers? Because they manage teams.

So they need to have very simple, actionable insights. It is the managers of managers, they do not need to see all the employee details, but they want to see the people performance of their managers to coach them. So for that we need to have some kind of infrastructure, to talk to these managers, to get feedback, to reinforce learning.

So if this is the case, if you really want to push it way beyond HR then we need to think back on what do we need today and what do we need to build today? And in product we call it refactoring, we do that every year. So every year we break down what we built at Crunchr and we rebuild it for speed, for stability, we test everything and that is the same product mindset as where people analytics need to be.

So this starts really, if we look at the tech stack, at its core. We need a really fast database, why fast database? Well we are fighting, as computer scientists call it, the big O. The big O means it is some kind of order of complexity and if you do a very simple regression analysis or you just want to have turnover by function over the course of three or four years that is expensive. It becomes even more expensive if you also want to do this by age, by gender, et cetera. So you need to have very strict and very stable architecture at the core then you need a very fast layer on top to access this data in a very consistent way to harmonise definition. Then on top, we have the authorisations because the GDPR police come every now and then and they say, hey, this HR person left the building what did this person do for the last five years?

So you really need to think product to bring as the skill. So that's one thing, product.

I always like to talk about tech and touch because it is not only technology or not only products but it is also the touch. You can have great insights but if you do not know what to do with it, then you may still make no impact.

So it is about educating the end user on how to use this, not how it works, but how to use it. This is very important. So again, think about the Target Interaction Model, then plan back on what kind of infrastructure you need, and again the wobbly stones do not work anymore they are too slow, you cannot authorise,

David Green: It is a risk as well

Dirk Jonker: It is a super risk. I mean, again, these two big companies, their reporting broke down and they can not trace back anymore what went wrong. It was untested and then they discovered that they were sending wrong results to the business. That is a consequence of just adding on top, on top, on top, and top and moving from one hype to the other.

We need to re factor our thoughts and think future.

David Green: So to summarise where you are thinking is think about the user, think about the user in the business. Whether it is a manager, a manager of manager or potentially even the employees themselves in some cases so they can get a return, they get some insight from providing their data.

Certainly I am seeing more and more people analytics teams now adding almost UX people within the team for some of the products that they are developing but also some of the third party products that they are bringing in as well. Because as you said, it is not just about giving it to the business partners, it is about actually democratising that data and giving it to the business and then allowing managers to take action on it.

Dirk Jonker: Exactly. Exactly. We did some research, two years ago, on what is actually going on right now because on one hand we have tons of HR processes that are management systems, they generate a lot of data and we have beautiful BI tooling generic. On the other hand, we have the business where they create value and drive change et cetera. If you have all these insights why doesn't HR have a seat at the table? You know, this horrendous discussion that has been going on for already 20 years.

So something in the middle is missing and what we felt was missing was basically that HR, you see a dashboard but if you do not trust the numbers, if you cannot play with the data to really make it real, then you will never use it. So this is a concept that we started to call guided analytics two years ago, where basically the first thing is to build trust and love with your data.

I see a turnover of top potential turnover of 12.6% in certain business units. Could be right. Could be wrong. Who are those people? Click, boom. You see the names.

Then you want to have it as some kind of a Rubik's cube, to see multiple perspectives of these data points and that leads basically to stories.

The stories lead to interactions between managers, HR and employees and eventually that drives change. So that is exactly why we see a lot of these UI experts entering HR, we see a lot of dev ops entering HR basically to manage the infrastructure that runs all these models. So the prototyping is done, the great potential showing is done, check in the box.

HR did fantastic work, but now it is really time to move on.

David Green: I think, as you said, really it is about giving people the data and actually not just giving people the data, but almost guiding them on what they should do with it. What is the data telling me and what is the action I can take. What is the outcome that is going to result from that. Then almost feeding it to managers rather than just expecting them to come to the conclusion themselves.

Dirk Jonker: It is about injecting these data points right into the workflow where needed and keeping it super simple. I see that you are wearing a watch, basically a step counter, HR would say walking is really good. You need to walk 10 kilometers per day, but here in the UK it is 6.21 miles per day.

Forget all of these definitions, you need to walk 10,000 steps per day and it keeps the doctor away. It is a little story that will stick in your mind, you will never forget. So we need to make things simple. We are over complicating things. A beautiful dashboard could just be three metrics that matter.

So it is much more than this elegant mathematics, which of course drove a lot of the art of the possible and now we need to quickly add different perspectives to build a product and to make it scaleable.

David Green: So one of the big themes I think around HR at the moment is this whole process of up-skilling.

I think we realise that HR needs to become more data literate and more digital. When we were talking last week, you had an interesting view on why you think a lot of organisations are approaching up skilling in the wrong way. What are they getting wrong and what should they do instead?

Dirk Jonker: Okay, so what I think is that there is a big need for up-skilling because most of the technology and most of the dashboards and the insight that we are pushing are way too complex. So, my mother has a MacBook, my mother has an Apple watch, she has never had any training because it's simple.

The second thing is that most HR people are interested in people analytics they just want to learn a bit about the narrative. But they did not choose the HR profession to become data professionals and as we mentioned before, the tech and touch, you needed technology to help you to prioritise the actions, but the touch is the human interaction needed to drive change. The latter is why most people chose the profession of HR. So technology needs to empower this interaction but not change people into a way they do not want to become.

So up-skilling is needed now because I think that technology and dashboarding is too complex.

David Green:  o make it less complex, make it easy to use and actually get over this kind of misnomer that all HR people need to become analysts because clearly they don’t, there are people with those skills that can do that. It is almost how can they take this data and how can it help them have better conversations with the people that are within the business or with employees. How can they become the translators?

Are you seeing more and more organisations looking at how they can make more of their HR professionals translators so that they can take the data and have the conversations?

Dirk Jonker: Yes but I also think that that is a very intermediate step because that is just another hype.

Also this translation is needed because the technology does not help us yet to take the final step. So If you really dream like 10 years ahead, how do I think the targeted interaction model should be or could be? That is basically that the HR professional has an app and it is being dispatched by some kind of centre of expertise, like hey David, we need to go to this unit because engagement is dropping, turnover is picking up. This is hurting the business. Please go check it out, talk to some people, by the way talk to these people and we think that these might be the reasons. You will have to have a human conversation and you will feed back the intelligence to the system.

So here we do not need to talk about translations it is just about how to use these insights in order to have a conversation. When you go to Japan and you rent a car, there is no manual. Here is the ignition, here is the book, here is the what? No, it is just three steps that would start the car.

Click, click, click great and off you go. We need to think way simpler.

David Green: Yes I think even I could drive that car. So, in our work at Insight222 we are working with a lot of organisations now who have people analytics teams or are trying to grow those people analytics teams further. One of the questions we get asked a lot is around technology.

Should we build technology or should we buy technology, particularly in the data aggregation or visualisation space. As someone who has been building technology in this area for over six years, I think it would be really interesting to get your advice for those companies who are thinking of building. and what approach they should consider.

Dirk Jonker: So my advice would be always start with building something yourself because what you need is velocity. You need to fast iterate to see what the business needs. You maybe want to try a smaller dashboard and to put it in front of managers and ask them if this is actually interesting to you? Would you actually use it?

As you start to add more features do a usage analytics to find out if people are actually using it and logging in. Then it is time to round up this MVP phase and reconsider the tech stack for the next three years. So if your ambition really is in the next three years to onboard loads of managers or to even open it up to all the HR professionals in the field, then you need to think product and you really need to think about what are the external requirements on authorisation, GDPR, anonymisation of leavers etc. What are the types of questions that our users are going to ask and what kinds of techniques are needed for this, how much computing power is needed and that basically describes the functional requirements.

If you have a lot of developers available and dev folks etc then you can always build it yourself. But what we have seen is that even though you have very big teams in HR, you are always dependent on IT. Typically HR does not get too much priority with IT and this is when you die basically of your own success.

So, and this is really where most companies are at right now, they have shown success, they get traction but they still are stuck with the HRIS data extraction, database visualisation etc.

David Green: The stones again.

Dirk Jonker: Yes the stones and that does not scale. So there comes a moment where you need to reconsider. Then some people say, wow this dashboard, this is also something we can build. I can also build a front end of the Google search engine, right? We can do that together in two hours, but what is behind and what is really needed to scale, that is quite difficult. It has taken us more than 150 man years to get what we currently have.

David Green: I suppose what happens is when you get to a certain size, the amount of time that gets taken up with doing this and maintaining it versus what you could be doing instead. Whereas, and I guess you have worked with companies who have done exactly that, created that MVP, and then you have helped them get to the next level, the scaling level, by bringing Crunchr onboard.

Dirk Jonker: Yes, exactly. Those are the best collaborations.

David Green: Because you have got that knowledge there, the team there, the desire there, they have already tested that desire within the business as well for having products like this. Rather than just trying to get budget for this at the start, as you have alluded to, it is very difficult.

Dirk Jonker: Exactly. Exactly.

David Green: So can you share an example of a client that you have worked with who has built their own technology and then not necessarily change their minds, but realises that to get to the next level they decided to partner with Crunchr?

Dirk Jonker: Yes, so I think about 80% of our clients are global companies. 30 to 40,000 employees and up. So let me just call out one because I am really proud of this collaboration. It is AkzoNobel they are around the world and they did a phenomenal job, by not only focusing on the technology but also a lot on the touch. So end result, hundreds of HR professionals are now using the product on a very frequent basis and that the engagement is picking up. Most of the clients are renewing the contracts, are doubling the contracts and this is really the proof that it starts to work. But again I do believe that companies need to start building something themselves. I am also really passionate for bigger companies that think they will build themselves further to help them.

We talk a lot at meet ups to give tips and give advice but it took 150 man years and woman years to build what we currently have. The typical company does not have 50 developers at their disposal for three years but neither does the business want to wait.

David Green: So part of your philosophy really is to help enable the community by sharing and then you are there if they need you at some point down the line.

Dirk Jonker: Exactly. Exactly. I teach people analytics at a couple of universities, which I really enjoy. I speak a lot for leadership teams, to inspire them on the art of the possible, but also bring them back to, hey, you have got to show ROI right now and make a plan.

That is really what I like to do.

David Green: I think that is good. One thing I have noticed about the people analytics space, after all the years I have been in it, is it is naturally quite collaborative. People are curious. They want to learn whatever stage of the people analytics journey they are on as individuals or their organisations are, and they want to share. Actually with the 150 man years that you have actually spent building this product there is a lot of stuff that you can share that helps people. I think that is very good that you are doing that.

Dirk Jonker: It is also the belief that we need to create this open ecosystem, since we are a data aggregator and analysis and reporting on top, we need to collaborate.

Our whole philosophy around people analytics is to show the bigger picture to our end users. So even though you are working in recruiting you need to know what the people want for whom you are actually optimising. What is the impact of your recruiting efforts? Are these people staying? What does the quality of hire?

So we need to collaborate with the data creators and also with a lot of universities because a lot of research, very fundamental research in computer science is happening right there.

David Green: So are there any other examples of how Crunchr has helped? We talked a lot about bringing people analytics to scale, have you got any other examples of how Crunchr has helped your clients to do that?

Dirk Jonker: Sure. One of the big themes nowadays is employee experience, is future of work. So let me give you two very small examples in each space. I really like the whole discussion around employee experience, but I think we are missing a really important element. We try to do employee experience analytics based on the data that we currently have in HR systems, while this shows the behaviour but we are missing the biggest part. What do people really want in their work and in their career? So we developed a little game to measure this. This is what we have deployed already more than 200 times for big companies around the world and it is asking one simple question. What is most important to you in your work and your career? Now we use this, together with extremely elegant mathematics, to build personas and to build an employee value proposition for total rewards for recruiting, employer branding, to attract those people that you need to have to win tomorrow.

David Green:  Can you segment that by job role, seniority and stuff like that?

Dirk Jonker: We could yes but we also take a completely different approach, we can show how people are different. We can also show how people are the same. This is where we use cluster analysis to segment people by groups on what they find important and that would become the personas that HR works with.

That has a result of, on average, a 22% higher return on investment on the annual total reward budgets so we are not spending money anymore on items that people do not really value.

The whole future of work, is also big. We mentioned financial institutions, there is a lot of compliance work coming their way with customer due diligence really struggling.

Also there again, we almost gamified the way we do workforce planning to make it accessible for people. We could talk a lot. I am sure we could talk for ages, including Ian, on what kind of mathematics we can use or whatever. But that is a useless discussion. We need to open it up to all the field HR and make it accessible to them.

So there we use government agencies, we use big banks and driver trees to start speaking business language and connecting them to the data in the Workday or whatever to build a plan and to show what is going on.

David Green: So let us look ahead. We are in the new decade now and the field, as we talked about right at the start, is moving really fast.

What excites you most about people analytics and what are your biggest concerns?

Dirk Jonker: What excites me most is that everybody is talking about people analytics and we have about a thousand very highly skilled professionals around the world that are doing projects and are open to share.

My biggest concern with people analytics is that we do not get out of this MVP phase and that the business leaders with the next recession are going to say, okay, so what was the real return on investment? How much are you spending in people analytics with all those conferences and all those people that you are hiring? Maybe we should bring that down a bit. So we are at a really critical point right now that we need to show value and bring it to production.

So stop talking about new hypes and make it real.

David Green: Yes I think that is fair. I think we see a lot of organisations that sometimes there is too much focus on getting the right skills in the team, the right data and everything else but needs to be more focused on what are the problems that we are trying to solve for the business and how do we measure the impacts of our insights and our interventions. That would then answer that question, that when the next recession comes we actually need more people analytics.

Dirk Jonker: Exactly. Everything that we need to do needs to have a direct line of sight with business outcomes. If we make sure of that then we should be fine.

David Green: So this leads onto the question that we are currently asking all the guests on the show. AI and automation, do you see them as an opportunity or a threat to HR? In your answer if you could describe the opportunities and any threats that you see as well?

Dirk Jonker: A big question. So let me try to be concise. I think there is no doubt AI and technology will definitely change the way that HR works, some people might be scared for the change but we will definitely overcome. It will bring so much to HR that the benefits will be amazing.

We both have phones, there is a junk mail folder on there. In the beginning when Microsoft said, hey, we are going to scan your emails and we are going to put some in a junk folder, we were like, Oh, that is scary because maybe good emails end up in the junk mail. Well you now take that for granted because it gives so much convenience. Whenever now you lose an email, you check your junk every now and then. It is okay, we accept it. But we keep talking about everything that can go wrong. So let's get over that. Let's look at the business value that we create.

The second thing is that we still talk a lot about math and we still talk about very specific topics while we need to grow everything in balance. Let me give you a quick example. Some companies are really stuck in this data discussion. We cannot do anything else because our data's not good.

Well, that is a chicken and the egg situation. If you do not show how bad the data is or what you can do with the data, the data will never improve, and to be honest the data will never be perfect. We can hire fantastic data scientists, but if your data is not up to par and they still need to do a lot of data engineering, you will get them onboard, excite them for three months and then they will leave.

So we need to grow everything in balance. So tons of opportunities, but HR needs to have the game plan right for the target interaction model otherwise they are just doing stuff. They are just running around like a chicken without a head. So time to take a step back to reconsider, think about the target interaction model, re factor everything that we built and then the future is extremely bright.

David Green:  And exciting as well.

Dirk Jonker: Super exciting

David Green: We have seen the growth of people analytics and clearly you can not do any of this fancy stuff around artificial intelligence without having good analytics to power it. Hopefully a good future for people analytics.

Dirk Jonker: Oh for sure.

David Green: Dirk it has been wonderful to have you as a guest on the show. How can listeners stay in touch with you and stay in touch with Crunchr?

Dirk Jonker: We have a website. We speak at the lots of meet ups and conferences. I do frequent leadership meetings where I just bring some inspiration, like yourself on, on the art of the possible and things will follow.

David Green: On social media where is the best way for people to follow you? Is it LinkedIn?

Dirk Jonker: LinkedIn is generally best. Still use a bit of Twitter, my Twitter name is quite easy it is @Datadirk

David Green: Datadirk. Okay. That is easy enough and then LinkedIn as well.

Dirk Jonker: Absolutely.

David Green: Dirk it has been great to have you on the show. Thank you for being a guest.

Dirk Jonker: Thank you.